I characterize the largest and to date most hydrocarbon-prolific passive-margin salt basins, including conjugate basins in the South Atlantic and Gulf of Mexico, as syn-exhumation salt. The traditional view is that single evaporite depocenters were subsequently divided by the accretion of oceanic crust (e.g. Evans, 1978; Duval et al., 1992; Davison, 1999; Karner & Gamboa, 2007; Pindell & Kennan, 2007; Mohriak et al., 2008). Others have argued that salt in some of these conjugate basins (such as the Kwanza and Campos basins) was never contiguous but separated by proto-spreading centres (e.g. Jackson et al., 2000; Quirk et al., 2013). Imbert & Philippe (2005) suggested that the Gulf of Mexico originally comprised one salt basin that was deposited over actively forming oceanic or proto-oceanic crust (characterized by SDRs) with no exposed spreading centre until ultimately splitting into two salt basins. I have defined syn-exhumation salt as predating final breakup, described as the localization of complete plate separation in a thermally and magmatically weakened spreading system of normal oceanic crust (Lavier & Manatschal, 2006; Péron-Pinvidic & Manatschal, 2009) or rupture immediately followed by seafloor spreading (Huismans & Beaumont, 2011). Whether the salt overlies thinned continental crust, exhumed mantle, volcanic crust with SDRs, or some combination is immaterial to the classification of the salt as syn-exhumation and pre-breakup.
Gulf of Mexico
There is broad consensus on the origin and timing of the Gulf of Mexico, with rift initiation in the Late Triassic and the onset of oceanic spreading by the latest Callovian to early Oxfordian as the Yucatan block moved away from North America, first to the SE and then rotating counter-clockwise about a pole located near western Cuba (e.g. Salvador, 1991; Pindell & Kennan, 2001, 2009; Kneller & Johnson, 2011). Evaporites formed at the end of this extensional history, primarily during the Callovian (Salvador, 1987). Hudec et al. (2013) suggested that spreading did not commence until the Kimmeridgian to Tithonian, so that there was a phase of crustal stretching between the end of salt deposition and the beginning of oceanic crust formation. In any case, the nature of transitional crust underlying depositional salt between the shoreline and abyssal plain is controversial due to deep burial and shallow salt. Recent publications include those that invoke: thinned continental crust (Roberts et al., 2005); proximal rifted continental crust and distal proto-oceanic crust (Pindell & Kennan, 2007); oceanic and proto-oceanic crust (Imbert & Philippe, 2005; Mickus et al., 2009); ultraslow-spreading lithosphere interpreted to comprise serpentinized mantle and/or areas of thin oceanic crust (Kneller & Johnson, 2011); and, for the northern Gulf of Mexico, hyperextended continental crust with a lateral transition from exhumed mantle in the west and centre to proto-oceanic crust in the east (Rowan et al., 2012b).
Despite this uncertainty, several lines of evidence show that the salt is syn-exhumation. First, the original extent of the evaporites is enormous, spanning the proximal onshore to the distal deepwater (Fig. 12). Second, deposition occurred just prior to final breakup and seafloor spreading. Third, the salt forms the upper portion of a thick sag basin in the southern Gulf of Mexico (Miranda et al., 2013). Finally, and in keeping with the previous point, the base salt is mostly unfaulted, whether it overlies SDRs (Fig. 13) or hyperextended crust and exhumed mantle (Fig. 14). In other words, salt was deposited after the bulk of brittle upper-crustal extension had ceased. A prominent exception is the so-called ‘stepup fault’ or ‘inner ramp’ (Hudec & Peel, 2010; Barker & Mukherjee, 2011; Pindell & Horn, 2012), but this marks the breakup edge of autochthonous salt, with more distal salt representing early allochthonous advance over oceanic crust (Fig. 14). In their most recent interpretation, Hudec et al. (2013) suggest that the inner ramp is not the limit of depositional salt sensu strictu but simply the boundary between oceanic crust and a zone of parautochthonous salt that was stretched over extended transitional crust after evaporite deposition (Fig. 12).
Figure 12. Map of Gulf of Mexico with present-day distribution of salt (modified from Hudec et al., 2013). Approximate locations of seismic profiles indicated by dashed ellipses.
Download figure to PowerPoint
Figure 13. Uninterpreted (a) and interpreted (b) versions of prestack depth-migrated 2-D seismic profile from the northeastern Gulf of Mexico. The thin salt is above a combination of more proximal rotated fault blocks of continental crust and more distal volcanic crust, with only minor offset of the base salt on several faults. A boundary between more and less reflectivity (deeper and shallower, respectively) is interpreted to represent the brittle-ductile transition between the upper and lower continental crust. Thin and thick purple lines are interpreted volcanic flows seaward-dipping reflectors (SDRs) and saucer-shaped intrusive sills, respectively. Vertical exaggeration 2 : 1, approximate location shown in Fig. 12. Data from SuperCache survey, courtesy of Dynamic Data Services.
Download figure to PowerPoint
Figure 14. Uninterpreted (a) and interpreted (b) versions of prestack depth-migrated 2-D seismic profile from the northwestern Gulf of Mexico. In (a), the red and black arrows highlight key basinward- and landward-dipping reflectors interpreted as a low-angle detachment fault and the Moho, respectively. In (b), the deep (autochthonous) salt level is above a combination of hyperextended continental crust, interpreted exhumed mantle and a possible thin sag sequence, with offset of the base salt on only one fault. The base salt ramps up basinward over the stepup fault and extends over oceanic crust as an allochthonous nappe, and a large, partly welded canopy is present at shallow levels. Rotated presalt fault blocks are visible above the low-angle crustal detachment fault (red) that can be mapped over an area of 200 × 100 km and that merges downward with a rising Moho. Vertical exaggeration 2 : 1, approximate location shown in Fig. 12. Data from SuperCache survey, courtesy of Dynamic Data Services.
Download figure to PowerPoint
Postsalt deformation in the northern Gulf of Mexico began immediately after salt deposition due to gravity gliding caused by basinward tilt of the margin as newly formed oceanic crust subsided (e.g. Peel et al., 1995; Rowan et al., 2004; Fort & Brun, 2012) or due to stretching of the salt and its cover during ongoing crustal extension (Hudec et al., 2013). Upper Jurassic to Lower Cretaceous extension was located near the proximal edge of the salt basin and was accommodated by shortening and nappe emplacement at the distal toe (see Fig. 14; Imbert & Philippe, 2005; Pindell & Kennan, 2007), a distance of up to greater than 500 km. This basinward translation set up Mesozoic depocenters and intervening areas of inflated salt that evolved into diapirs. Extension, translation, and contraction continued during the Cenozoic but during this time was driven more by progradational loading and consequent gravity spreading (e.g. Worrall & Snelson, 1989; Peel et al., 1995; Rowan et al., 2004, 2012a), although some gravity gliding was caused by proximal uplift of the margin (Jackson et al., 2011; Dooley et al., 2013). Allochthonous salt is almost ubiquitous, ranging from individual sheets less than 100 km2 in size to enormous, sometimes multi-tiered canopies covering more than 80 000 km2 (e.g. Diegel et al., 1995; Pilcher et al., 2011). Cenozoic gravitational failure was partitioned in spatially and temporally complex patterns between the autochthonous and allochthonous levels, depending on large part on the size, position, and base-salt relief of the canopies (Rowan & Inman, 2006; Rowan et al., 2009).
The central South Atlantic (Fig. 15) contains salt basins on both the western, Brazilian margin (Santos, Campos, Espirito Santo, Cumuruxatiba, Jequitinhonha, Camamu, Sergipe-Alagoas) and eastern, African margin (Namibe, Benguela, Kwanza, Lower Congo, South Gabon, Rio Muni). Somewhat surprisingly, there is even less consensus on the crustal architecture and timing than in the Gulf of Mexico. Estimates of the beginning of rifting range from the Berriasian (e.g. Karner et al., 2003; Mohriak et al., 2008; Lentini et al., 2010; Unternehr et al., 2010) to the late Valanginian (Quirk et al., 2013) to as late as the early Barremian (Meisling et al., 2001), and final breakup is thought to have occurred either during the Barremian (Jackson et al., 2000; Marton et al., 2000; Quirk et al., 2013) or near the Aptian-Albian boundary (Meisling et al., 2001; Karner et al., 2003; Karner & Gamboa, 2007; Torsvik et al., 2009; Unternehr et al., 2010; Blaich et al., 2011; Mohriak & LeRoy, 2013).
Figure 15. Present-day Aptian salt thickness in the South Atlantic shown on a map reconstruction immediately prior to oceanic breakup (modified from Lentini et al., 2010); original salt distribution would have been broadly similar, although there has been bulk movement of salt basinward. ‘COB’ indicates interpreted continental-oceanic boundary; salt basins in italics. Dashed ellipses indicate the approximate locations of the profiles shown in Figs 16 and 17.
Download figure to PowerPoint
The underlying cause for the discrepancy in breakup age centres on the origin of the so-called sag basins (or presalt wedges). These are broad, mostly unfaulted packages of Barremian to Aptian strata up to 7 km thick (e.g. Henry et al., 1995, 2004; Karner & Gamboa, 2007) between salt and a deeper unconformity that overlies rifted crust. In older interpretations, the sag basin and salt thin both landward and basinward, but newer data suggest that the distal limit is more abrupt where the basin steps up onto an outer high of either oceanic or continental affinity (Lentini et al., 2010; Unternehr et al., 2010; Zalan et al., 2011; Davison et al., 2012; Quirk et al., 2013). Some consider the sag basin and evaporites to represent the earliest postrift strata, deposited above the breakup unconformity and thus after plate separation (e.g. Jackson et al., 2000; Marton et al., 2000; Quirk et al., 2013), whereas others interpret the sag and salt as part of the synrift section that predates breakup (e.g. Karner et al., 2003; Karner & Gamboa, 2007; Mohriak et al., 2008; Unternehr et al., 2010; Blaich et al., 2011).
The evaporites themselves have variable age estimates ranging from as early as 124 Ma to as late as 110 Ma (Davison, 2007; Karner & Gamboa, 2007; Mohriak & LeRoy, 2013; Quirk et al., 2013), so were either pre- or post-breakup depending on the interpretation. Thus, the salt was deposited either in a single depocenter that was subsequently separated into the Brazilian and African basins (e.g. Karner & Gamboa, 2007; Mohriak et al., 2008; Torsvik et al., 2009; Lentini et al., 2010) or in separate basins on either side of a subaerial spreading centre (e.g., Jackson et al., 2000; Marton et al., 2000; Davison et al., 2012; Quirk et al., 2013). Furthermore, there is disagreement on what type of crust underlies the most distal autochthonous salt: thin continental crust, with or without a significant proportion of magmatic rocks (e.g. Mohriak et al., 2008; Lentini et al., 2010; Mohriak & LeRoy, 2013); volcanic crust, whether oceanic or consisting of SDRs (e.g., Jackson et al., 2000; Marton et al., 2000; Torsvik et al., 2009; Quirk et al., 2013); exhumed subcontinental mantle (e.g. Unternehr et al., 2010; Zalan et al., 2011); or exhumed lower continental crust (Sibuet & Tucholke, 2013).
Taking into consideration all the various models cited above, the newest seismic data with the best deep images, and analogies to other passive margins such as the northern Gulf of Mexico, I find interpretations similar to those by Unternehr et al. (2010) and Zalan et al. (2011) to be the most compelling. What is beyond question is that evaporite deposition postdated the bulk of crustal thinning, so that there is generally only local and minor offset of the base salt (Fig. 16). There are certainly some large faults, especially on the Brazilian margin, but much of the base-salt relief is due to some combination of differential compaction and drape above older faults, reactivation due to loading by the excess salt, and post-breakup tectonics (Davison, 2007; Mohriak et al., 2008; Davison et al., 2012). One prominent exception is the step up of the base salt onto either oceanic crust (Fig. 16a) or exhumed mantle (Fig. 16b). In any case, the salt was most likely deposited in a single basin that was subsequently separated during final breakup and accretion of normal oceanic crust. The salt and presalt sag basin are underlain by hyperextended continental crust and, more distally, some combination of exhumed mantle (as depicted in Fig. 16) and SDRs.
Figure 16. Cross-sections from interpreted seismic data in the South Atlantic: (a) profile from the Lower Congo Basin, Angola (modified from Unternehr et al., 2010); (b) profile from the Campos Basin, Brazil (modified from Zalan et al., 2011). Both show salt with relatively low-relief bases over sag basins and hyperextended continental crust, salt concentrated in distal areas, allochthonous salt ramping up over stepup faults and emplaced over oceanic crust (a) or mantle (b), and exhumed subcontinental mantle that is infiltrated and/or serpentinized (diagonal lines). Vertical exaggeration 2 : 1.
Download figure to PowerPoint
Gravitational failure may have begun while salt was still being deposited (Davison et al., 2012; Quirk et al., 2012) but was certainly ongoing during and after deposition of the overlying Albian carbonates. Basinward movement was in some cases convergent or divergent (Cobbold & Szatmari, 1991). Proximal extension is recorded by a combination of basinward-dipping and counterregional faults, including the well known Cabo Frio Fault or so-called Albian gap in the Santos Basin, often separating the oldest suprasalt section into extensional rafts (e.g. Duval et al., 1992; Demercian et al., 1993; Mohriak et al., 1995; Marton et al., 2000; Fort et al., 2004). Translation over ramps in the base salt, for example where the salt drapes over the Atlantic Hinge Zone in offshore Angola, created a series of landward-shifting depocenters (Hudec & Jackson, 2004; Rowan et al., 2004; Jackson & Hudec, 2005). Distal contraction is characterized by folds, thrusted folds, squeezed diapirs, squeezed and inflated salt massifs, and thrust emplacement of allochthonous salt out over oceanic crust (e.g. Demercian et al., 1993; Cobbold et al., 1995; Marton et al., 2000; Rowan et al., 2004; Hudec & Jackson, 2004; Brun & Fort, 2004; Fiduk & Rowan, 2012). Allochthonous salt is present, but not to the same extent as in the Gulf of Mexico. Moreover, the canopies that do exist did not serve as secondary décollements for gravitational failure as they did in the Northern Gulf of Mexico.